Fleet Operations.
Eighteen vehicles. Four dispatchers. COGITA TMS, Webfleet telematics, and DATEV accounting connected through a single definitional layer. No system replaced.
An owner-managed road freight operator in Lower Saxony — eighteen vehicles, four dispatchers, approximately €4.5 million annual revenue — running three separate systems with no shared definition of what a fleet event means. COGITA TMS managed orders. Webfleet tracked vehicles. DATEV processed invoices. None of the three shared a representation of what happened when a driver confirmed a delivery. A regulatory inquiry from the Landesbeauftragter für den Datenschutz Niedersachsen exposed the absence of documented legal basis for GPS data the company had been collecting for years. Stathon deployed at Forge Tier: the decision cycle is now continuous, status-check calls have fallen by 45–55%, approximately 120 dispatcher hours per month have been freed, and a full DSGVO compliance framework — DSFA, RBAC, retention policy, audit trail — is in production.
The operation.
An owner-managed road freight company operating eighteen mixed vehicles — 7.5 to 18 tonne, MAN and Mercedes-Benz — across Lower Saxony and neighboring federal states. Regional distribution: retail deliveries, building materials, industrial components. Approximately €4.5 million annual revenue. No internal IT team.
Four dispatchers managed daily tour planning through COGITA, the Weber Data Service TMS. Webfleet provided telematics — GPS positions, driving behavior scores, fuel data — for all eighteen vehicles. DATEV handled accounting through a nightly batch export from COGITA's invoicing module. Scheduling lived in Excel. The three systems were individually functional. They shared no operational language.
The dispatch team was experienced and capable. One senior dispatcher carried the mental model of the entire fleet — vehicle states, driver preferences, reliable customers, problematic routes. The operation worked because she made it work. That is not a criticism. It is a description of institutional fragility disguised as operational stability.
Owner-managed, ~€4.5M annual revenue. No internal IT team.
Three systems. No shared event.
COGITA, Webfleet, and DATEV each captured operationally relevant data. The problem was not access to data — the problem was that none of the three systems shared a definition of what a fleet event means. A driver arriving at a delivery address was simultaneously a position update in Webfleet, an order status change in COGITA, and an invoicing trigger that would eventually reach DATEV. These were three separate representations of one event. No system held the unified representation.
The operational consequence was temporal fragmentation. COGITA data arrived in 15-minute extracts. Webfleet updated in near-real-time. DATEV ran nightly. The dispatch team operated on a plan that was accurate in the morning and increasingly fictional by afternoon. By the time the three data streams could be reconciled — manually, each morning — the previous day was already settled and the new day had begun without a coherent picture of the fleet.
The dispatch team was not working from bad data. They were working from yesterday's data, continuously. The absence of a unified fleet event definition meant the operational reality and the system of record were always out of phase.
Stathon engagement assessment
A DSGVO audit inquiry from the Landesbeauftragter für den Datenschutz Niedersachsen, prompted by a former employee's data access request, revealed the company could not demonstrate a compliant legal basis for the GPS data it held.
No unified definition. COGITA defined orders. Webfleet defined positions. DATEV defined transactions. The moment a driver confirmed a delivery — simultaneously updating order status, generating an invoicing trigger, and producing GPS-timestamped data subject to DSGVO — had no unified representation.
The dispatch picture in COGITA diverged from reality throughout each day. By afternoon, the plan no longer matched the fleet. Reconciliation happened the following morning, when the cycle restarted.
No formal classification existed for which fleet data elements were personal data under DSGVO. Driver GPS positions, route histories, and behavior scores were stored without documented legal basis, impact assessment, or retention limits.
COGITA data arrived in 15-minute extracts. Webfleet updated in near-real-time. DATEV ran nightly. The event stream operated at three different speeds with no reconciliation mechanism.
Previous approaches
A previous consultant had proposed replacing COGITA with a larger integrated suite — several hundred thousand euros, multi-year migration. The managing director declined, correctly identifying that the tools worked and the problem was not the tools. The DSGVO exposure had been flagged internally as a compliance gap requiring legal review. It had not been recognized as the surface manifestation of a structural definitional absence that also governed the operational problem. Both problems shared a common root.
Three phases.
Forge-tier deployment. Four modules — Arché, Core, Athena, Aegis — deployed in phased sequence. All three source systems remained in place, unchanged. No data was migrated. No system was replaced.
Definition & Integration Spine
Weeks 1–6The first work was not integration — it was classification. The legal basis for GPS data had never been documented. Making the definitional structure explicit was the prerequisite for both the compliance framework and the operational dispatch view.
Stathon deployment record
Consolidated Dispatch View
Months 2–4Unified view: vehicle position from Webfleet, tour progress from COGITA, delivery confirmations as they occur, working-time alerts where driver hours approach regulatory limits. The dispatch picture is maintained continuously rather than reconstructed each morning.
The senior dispatcher adopted within the first week. Mid-level dispatchers took approximately three weeks. The junior dispatcher — the most recent hire, with no embedded habits to override — adopted fastest of all.
Geofence matching against COGITA stop coordinates achieves approximately 91% automated detection after recalibration. Initial rate was 82%; a tuning cycle adjusted geofence radius and dwell-time threshold for industrial delivery sites with large premises.
Route Deviation Scoring
Current · OngoingAthena route deviation scoring is operating in advisory mode. Weekly management review rather than real-time alerts — the false positive rate is being tuned before dispatcher-visible alerting is activated. Three deviation patterns have been identified: a loading window mismatch on a recurring retail delivery route, B6 corridor congestion displacing scheduled mid-morning arrivals, and a driver route that consistently outperforms the COGITA tour plan and has since been adopted as the default.
False positive rate has moved from approximately 30% at initial deployment to approximately 18% after two tuning cycles. Target for dispatcher-visible alerting is below 15%. One additional tuning cycle is planned.
One Athena-flagged deviation was reviewed and found to be superior to the planned route. The driver’s approach has since been encoded in COGITA as the default tour for that stop sequence. The system identified an improvement the plan had not captured.
Route deviation analysis operates at the tour level, not the individual driver level. Athena outputs flag route-sequence anomalies for management review. Attribution to specific drivers is not part of the scoring model — consistent with BDSG §26 constraints on performance monitoring.
What changed.
Measured impact from Phase 1 and Phase 2 in production. Route deviation scoring (Phase 3) is in advisory pilot; operational results pending alert activation.
This is not a faster dispatch screen. It is a structural change in the temporal relationship between the dispatch team and the fleet's operational state. What was previously reconstructed each morning from yesterday's data is now maintained continuously. The decision cycle did not accelerate — it was replaced by a fundamentally different operational cadence.
Definitional infrastructure is not reserved for enterprise-scale deployments. The operational physics are identical: fragmented data produces fragmented decisions; coherent data produces coherent decisions. Eighteen trucks generate the same structural problem as six hundred. The number of events per day changes. The nature of the problem does not.
What this case revealed.
Controlled inefficiency as steady state.
The company was not in crisis. It was in a state of controlled inefficiency sustained by institutional memory — specifically, one senior dispatcher’s mental model of the entire fleet. The system worked because she made it work. This is the most common condition we encounter: not failure, but fragility disguised as stability.
DSGVO as engagement catalyst.
The regulatory near-miss opened the engagement. But the data governance problem and the operational fragmentation problem shared a common root: the absence of a structural layer that defined what data existed, what it meant, and what rules governed its lifecycle. Solving one required solving both.
No system was the problem.
A previous consultant proposed replacing COGITA with a larger integrated suite — several hundred thousand euros, multi-year migration. The managing director declined, correctly identifying that the tools worked and the problem was not the tools. Definitional infrastructure operates above existing systems. It does not require an organisation to abandon what it built.
The integration is not visible from the surface. Its absence would be.
Stathon deployment conclusion
Forward roadmap.
Route deviation active alerting
Transition from weekly management review to real-time dispatcher-visible advisory alerts. Requires false positive rate stabilization below 15% — currently at approximately 18% after two tuning cycles.
Idle-time anomaly detection
Flags vehicles with statistically unusual idle patterns for operational review. Addresses fuel waste and schedule adherence without individual driver surveillance — aggregated at fleet level, not attributed to specific drivers.
Working-time threshold prediction
Probabilistic model estimating whether a driver will approach EU driving-time regulation limits before tour completion. Enables proactive rescheduling rather than reactive compliance.
Deployment record.
Stathon · Definitional Infrastructure Company. Client identity withheld by agreement. Deployment metrics reflect production conditions as of March 2026.